See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/251711712

Technologies for Gas Turbine Power Generation with CO2 Mitigation

Article  in  Energy Procedia · December 2011

DOI: 10.1016/j.egypro.2011.02.069

CITATIONS READS

12 601

5 authors, including:

Peter Jansohn Timothy Griffin

Paul Scherrer Institut University of Applied Sciences and Arts Northwestern Switzerland
54 PUBLICATIONS   869 CITATIONS    32 PUBLICATIONS   1,342 CITATIONS   

SEE PROFILE SEE PROFILE

François Maréchal Frank Jörg Clemens

École Polytechnique Fédérale de Lausanne Empa - Swiss Federal Laboratories for Materials Science and Technology
562 PUBLICATIONS   11,477 CITATIONS    192 PUBLICATIONS   3,070 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Strategic energy planning under uncertainty View project

Waste-to-energy multi-feedstock systems as an integrated waste management strategy in Colombia - UrbanWtE View project

All content following this page was uploaded by Peter Jansohn on 17 September 2015.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com
Energy
Energy Procedia 4 (2011) 1901–1908
Procedia
Energy Procedia 00 (2010) 000–000
www.elsevier.com/locate/procedia
www.elsevier.com/locate/XXX

GHGT-10

Technologies for Gas Turbine Power Generation

with CO2 Mitigation

Peter Jansohna1*, Timothy Griffinb, Ioannis Mantzarasa,

Francois Marechalc, Frank Clemensd
a
Paul Scherrer Institute (PSI), CH-5232 Villigen PSI, Switzerland
b
University of Applied Sciences (FHNW), CH-5210 Windisch, Switzerland
c
Swiss Federal Institute of Technology (EPFL), CH-1015 Lausanne, Switzerland
d
Swiss Federal Institute of Material Science & Technology (EMPA), CH-8600 Dübendorf, Switzerland

Elsevier use only: Received date here; revised date here; accepted date here

Abstract

A review of the state of the art gas turbine power plants designs with CO2 capture was performed to set up a list of
technologies to be considered in a conceptual process superstructure. It has been found that the efficiency and
economics of carbon dioxide capture in gas turbine combined cycle power plants can be remarkably improved by
introducing Flue Gas Recirculation (FGR) so as to increase the CO2 concentration in the flue gas and to reduce the
volume of the flue gas treated in a CO2 capture plant. Thus, this process was chosen as the main focus of the thermo-
economic modeling and the combustion system related experimental testing.
Combustion studies were performed to quantify the impact of reduced oxygen content (caused by flue gas
recirculation) on combustion stability and emission characteristics (NOx, CO). Tests were performed with methane,
methane/ethane (simulated natural gas), methane/hydrogen and natural gas (as distributed in Switzerland) at
different FGR ratios. For all conditions the addition of ethane or hydrogen shows beneficial effects (restoration of
flame reactivity) on flame stability and CO emission. Using PSI’s high pressure test rig, it has been shown, that by
redirecting up to 30% of exhaust gas volume back to the combustor, the CO2 concentration in the exhaust can be
increased to 8% vol. while keeping the flame temperature constant (Tad = 1750 K) and meeting the emission
requirements (CO, NOx < 25ppm @ 15% O2).
Benchmark tests of catalytic partial oxidation reactors were performed to determine whether sufficient H2 could be
produced in situ by reaction of a slip stream of natural gas with extracted compressor oxidant.
The work confirmed that FGR has the potential to significantly reduce the energy penalty connected with post-
combustion carbon capture techniques and still meet the requirements on NOx and CO emissions.
⃝
©
c 2010
2011 Elsevier
PublishedLtd.
by All rightsLtd.
Elsevier Open access under CC BY-NC-ND license.
reserved

Keywords: fluegas recirculation, gas turbine, power generation, thermo-economic modeling, emission (NOx, CO)

* Peter Jansohn. Tel.: +41-56 310 2871; fax: +41-56 310 4416.
E-mail address: peter.jansohn@psi.ch.

doi:10.1016/j.egypro.2011.02.069
1902 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908

2 Author name / Energy Procedia 00 (2010) 000–000

1. Introduction & Approach

The focus of the project activities is concerned with technologies for large (400 MW)
combined cycle plants with CO2 capture. In terms of the overall process the target is to achieve a
CO2 avoidance cost of less than 25 € per ton CO2 for gas turbines.
Although the literature is filled with many creative gas turbine cycle options for CO2 removal,
many of these cycle variations require significant modifications to the key turbomachinery
components in order to handle the modified fluid compositions. By simply applying flue gas
recirculation (FGR) in order to facilitate CO2 removal via increased CO2 concentrations in the
exhaust gas, most of these complications for the turbomachinery components are avoided or
minimized. The gas turbine process evaluation in this paper thus focused on studies on the effect
of exhaust gas recycle upon cycle performance.
The main effort was directed towards process design, using thermo-economic modeling and
optimization. The goal is to develop a process superstructure including different options for a gas
turbine power plant with CO2 capture capability. By applying a consistent methodology that is
based on thermo-economic modeling and energy integration techniques using multi-objective
algorithms, different options have been analyzed, optimized and compared systematically in
order to reveal which are most promising. Based on the developed models, the impact of CO2
recirculation on compressors, turbines, combustion, CO2 capture and the steam network has been
assessed.
With flue gas recirculation major modifications though are required for the combustion
section as the combustion properties of highly diluted fuel mixtures can change significantly.
Due to the importance of the recirculation rate on the overall performance of the CO2 removal
process combustion studies are required to define the maximum achievable recirculation rate.
Combustion stability concerns dictate hereby the minimum amount of oxygen excess
concentration allowable. The emissions of other combustion pollutants must remain at current
levels (NOx < 25 ppm, CO < 10 ppm).
The questions addressed in the experimental studies have been related to issues such as the
impact of a low O2 excess and high level of inert gas on combustion stability and emissions, as
well as if and how combustion stability can be improved by adding highly reactive species (H2-
rich fuel gases) into the fuel mixture (preferably based on an integrated production of syngas or
hydrogen).
High pressure combustion studies have been performed to determine the impact of flue gas
recirculation on the lean blow out limit and emission of CO, NOx for methane and simulated
natural gas mixtures. For a given fuel injection geometry, inlet temperature and pressure,
combustion stability limits have been determined for conditions corresponding to up to 50% flue
gas recirculation. The target has been to have stable combustion at less than 2 %vol. of excess
oxygen concentration.
The impact of hydrogen doping and hydrogen injection into flame stabilization zones of a
combustor has been measured. The target was to achieve similar combustion stability (quantified
by the residence time necessary for complete combustion) for various degrees of flue gas
recirculation. The impact of hydrogen injection on minimum required level of excess oxygen has
been identified.
The performance of pilot scale catalytic partial oxidation reactors has been determined as a
function of methane/oxygen ratio, pressure, temperature, and residence time. These reactors are
designed for integration into a fuel injector of a gas turbine combustor.

.
 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908 1903

Author name / Energy Procedia 00 (2010) 000–000 3

2. CO emission characteristics for CH4/air mixtures with dilution simulating FGR

The goal of this experimental effort was to determine whether NOx and CO emissions could
be maintained within allowable limits when FGR is applied and whether the operation range is
still compatible with requirements (mainly firing temperatures) imposed by the gas turbine
system. A series of measurements was taken for every FGR rate, covering a wide range of
fuel/air ratios limited only by the maximum allowable exhaust gas temperature (1400°C; as
measured) and by excessively high CO emissions near the lean flame blowoff limit (high excess
air ratio). The limits for the emission values of CO and NOx were chosen to be 25 ppm
(normalized to 15% O2), which is still the widely accepted limit for large stationary power
plants.
Figure 1 shows a typical example for the CO emission trends observed for varying air excess
ratios (i.e. flame temperatures) when simulated flue gas recirculation is applied. A pronounced
minimum in the CO emission can be observed for most of the chosen pressure conditions
(indicated by the dashed black line at λ =1.85). Towards even leaner fuel/air ratios (λ >1.85) the
steep rise in the CO emission is a result of increased dilution and decreased flame temperature
which causes the flame to approach its lean blow out limit. The decreasing flame temperature
slows down the chemical kinetic reaction rates required for the complete conversion of the fuel
components including CO (as the most important product of partial oxidation of hydrocarbon
species). For less lean conditions (λ <1.85) a steep increase of the CO emissions is likewise
observed, but, in this case, it is coupled with a certain pressure dependence (most significantly
seen for the highest pressure data, p = 20bar). For these fuel/ratios and the related (high) flame
temperatures high CO chemical equilibrium values (due to dissociation of carbon dioxide) are
the determining factor for the CO emissions. High pressure conditions are thus confirmed to be
favorable for low CO emissions, especially for high FGR.

Figure 1: CO emission in dependence of the fuel/air ratio for three different pressure levels (FGR = 10%)

In Figure 2 the results for CO emission for three different FGR rates at constant pressure (20
bar) are shown. This was the highest pressure at which data could be acquired so far (limitation
given by the air supply infrastructure), and matches typical pressure conditions of modern gas
turbine engines. At this high operating pressure the introduction of even high amounts (30%) of
1904 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908

4 Author name / Energy Procedia 00 (2010) 000–000

FGR did not change the CO emission level significantly, as long as the decreased reactivity due
to high FGR could be counteracted by running the flame less lean (keeping the flame
temperature up at levels of around 1750 K). This compensation measure does not work however
for an FGR rate of 40%. For these conditions it is not possible to achieve CO emission levels
below 30 ppm CO for the given combustor geometry and flow rates.


Figure 2: CO emission characteristics in dependence of the fuel/air ratio for three different FGR rates (p = 20bar).
Green/red markers indicate flame temperatures of 1750K/1850K, respectively.

To restore the original flame reactivity (judged by the flame location and the CO emission
without FGR) hydrogen was added to the fuel mixture either as pure hydrogen or as part of a
syngas mixture (containing 60 % H2 and 40 % CO by volume)in order to move the flame back to
its original location.

Figure 3: Required level of H2 to restore reactivity as a function of O2 vol % in the exhaust gas. Tests performed at 8
bar and 1750 K adiabatic flame temperature. Hydrogen added to the fuel via pure hydrogen (blue line) or syngas
60% H2 and 40% CO (red line).
 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908 1905

Author name / Energy Procedia 00 (2010) 000–000 5

The overall amount of hydrogen necessary to compensate for the reduction in reactivity due to
the depletion of oxygen was measured for a variety of conditions for Swiss natural gas, Figure 3.
It was shown that the inlet temperature played a minor role, but also that less hydrogen was
required if syngas was added, indicating that CO also plays a stabilizing role.
The concept of adding hydrogen to restore the reactivity of flames in case of FGR has been
followed also in the conceptual studies for process modeling. It has been evaluated to what
extent an in-situ generation of hydrogen-rich syngas (via steam reforming or catalytic partial
oxidation) influences the performance of a combined cycle plant with CO2 capture.

3. Thermo-economic modeling of novel gas turbine cycles

The conceptual superstructure for future gas turbine power plants designs represented in
Figure 4 includes pre-combustion, oxy-combustion and post-combustion options for CO2
capture, as for example a natural gas gas-turbine combined cycle plant (NGCC) with exhaust gas
recirculation, an advanced zero emission gas turbine power plant (AZEP) and a catalytic partial
oxidation (CPO) power plant.

Figure 4: Conceptual superstructure of gas turbine power plants with CO2 capture

This project is focused on power plants with post-combustion CO2 capture, consisting of a gas
turbine with flue gas recirculation, a CO2 capture unit (chemical absorption with amines) and a
steam network. This system can be represented in a more classical scheme in Fig. 5.
1906 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908

6 Author name / Energy Procedia 00 (2010) 000–000

Figure 5: Process scheme

Figure 6: Composite curve: Steam network integration with the gas turbine and the CO2 capture unit
without FGR and including district heating potential.

Compared to the base case without FGR and without CO2 mitigation, capture of around 90% of the CO2
emissions induces a thermodynamic process efficiency penalty in the range of 10 points if one considers
a preliminary simple, non-optimized steam network as illustrated in Figure 6. This representation is widely
used in energy integration. One of its main advantages is that the surface between the process (in red)
and the steam network (in blue) is an indicator of the losses.
 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908 1907

Author name / Energy Procedia 00 (2010) 000–000 7

Figure 7 shows the net electricity produced demonstrating that the increase of power
consumption due to syngas production can be more than compensated by the increase of power
produced by turbines (with the assumption of constant turbine isentropic efficiency).

Figure 7: Influence of FGR on the net electricity generation

The efficiency appears to be optimal at 40 % recirculation as shown in Figure 8. However,

variation is smaller than 1 %. Two points have to be taken in consideration. First a fix steam
network design has been used for every points of the sensitivity analysis, whereas, it must be
optimized for each configuration. Moreover, the isentropic efficiency has been assumed to be
constant. This underlines the importance of simulation of turbomachinery with different fluids.

Figure 8: Efficiency variation for different amounts of FGR

 1908 P. Jansohn et al. / Energy Procedia 4 (2011) 1901–1908

8 Author name / Energy Procedia 00 (2010) 000–000

Summary & Conclusions

Combustion testing has determined the impact of both oxygen content and fuel composition
on combustion stability, for gas turbine processes with flue gas recirculation (FGR). These
results have been incorporated into the thermo-economic process calculations.
A pronounced minimum in the CO emission can be observed for most of the chosen pressure
conditions with high pressure conditions (20bar) being confirmed to be favourable for low CO
emissions, especially for high FGR. This highlights a specifically high risk of elevated CO
emission if a gas turbine engine is run at part load condition (i.e. at low/moderate pressure
levels).
To restore the original flame reactivity (judged by the flame location and the CO emission
without FGR) hydrogen addition (either as pure hydrogen or as part of a syngas mixture
(containing 60 % H2 and 40 % CO by volume) can be very effectively applied.
In summary it can concluded that for adiabatic flame temperatures of around 1750 K - which
is a typical operating range for gas turbine combustors - both emission species of concern (NOx
and CO) can be kept safely below 25 ppm @15% O2 with FGR rates up to 30%.

The efficiency and economics of carbon dioxide capture in gas turbine combined cycle power
plants can potentially be improved by introducing Flue Gas Recirculation (FGR) so as to
increase the CO2 concentration in the flue gas and to reduce the volume of the flue gas treated in
the CO2 capture plant. Thus, this process was chosen as the main focus of the thermal-economic
modeling performed.
Different process configurations were investigated in order to study the impact of FGR on
combustion and CO2 capture. The impact of hydrogen injection to stabilize the combustion has
been studied by considering several ways of integrating the hydrogen or syngas production
processes. The results show that the detrimental effect of FGR on the combustion performance
(flame stability, CO emissions) can be counteracted with the addition of reactive species (such as
hydrogen/syngas) and that this measure does not compromise the efficiency of the power
generation process. Economic advantages (reduced CO2 avoidance cost) thus cannot be expected
via a reduced energy penalty in the operation of a post-combustion capture process but only via
reduced capital cost due to a more compact component design.

Acknowledgements
The authors want to acknowledge the financial contribution provided by SwissElectricResearch and the
Competence Center for Energy & Mobility (CCEM) in Switzerland.

The work performed was a significant team effort with major contributions from additional team members
at EPFL (L. Girardin, M. Dubois, L. Tock, F. Marechal),
at FHNW (S. Reimer, D. Winkler, T. Griffin),
at EMPA (M. Salehi, F. Clemens)
and at PSI (S. Karagiannidis, I. Mantzaras, M. Matuszweski, P. Jansohn)

View publication stats